Acute Complications Associated with Bedside Placement of Feeding TubesUniversity of Illinois College of Medicine at Rockford, Rockford, Illinois Correspondence: William N. Baskin, MD, FACP, FACG, University of Illinois College of Medicine at Rockford, 401 Roxbury Road, Rockford, IL 61107-5078. Electronic mail may be sent to drbaskin{at}rockfordgi.com. Several types of feeding tubes can be placed at a patient's bedside; examples include nasogastric, nasointestinal, gastrostomy, and jejunostomy tubes. Nasoenteral tubes can be placed blindly at bedside or with the assistance of placement devices. Nasoenteric tubes can also be placed via fluoroscopy and endoscopy. Gastrostomy and jejunostomy tubes can be placed using endoscopic techniques. This paper will describe the indications and contraindications for different types of tubes that can be placed at the bedside and complications associated with tube placement. Complications associated with nasoenteral tubes include inadvertent malpositioning of the tube, epistaxis, sinusitis, inadvertent tube removal, tube clogging, tube-feeding-associated diarrhea, and aspiration pneumonia. Complications from percutaneous gastrostomy and jejunostomy tube placements include procedure-related mishaps, site infection, leakage, buried bumper syndrome, tube malfunction, and inadvertent removal. These complications will be reviewed, along with a discussion of incidence, cause, treatment, and prevention approaches. Enteral nutrition has been widely used for nutrition support in critically ill patients. Numerous clinical and animal studies have supported the concept that early enteral nutrition may prevent bacterial translocation and maintain gut integrity and may decrease the known complications of sepsis and multiple-organ failure syndrome in the intensive care unit.1 The timing of enteral nutrition appears to be important, as early feeding may prevent a number of critical care complications, including prolonged length of stay and increased overall infections.2–5 Controversies remain, however, as to what is the preferred route of enteral access, whether it should be nasogastric (NG) or nasojejunal (NJ), and how soon, if ever, should percutaneous endoscopic access be carried out for long-term enteral feeding.6 Despite the above-described potential benefits, numerous complications have been reported for these various access procedures. Some of these complications may be preventable; all of them need to be managed. The purpose of this paper is to review the various complications associated with insertion and management of enteral feeding devices, with an emphasis on the incidence of the complication, presumed causes, management suggestions for the complication, and any potential preventive measures currently available.
Nasoenteric tubes are commonly used for short-term enteral nutrition in critically ill and general-floor hospitalized patients. Complications of nasoenteric tubes frequently include inadvertent malposition, epistaxis, sinusitis, inadvertent tube removal, tube clogging, tube-feeding-associated diarrhea, and aspiration pneumonia.
Inadvertent Malposition From Blind Bedside Placement An additional report by McWey et al8 described 14 misplacements out of 1100 nasoenteric intubations over an 18-month period, or a 1.3% rate of malpositioning. Thirteen out of the 14 had a pleura/pulmonary complication from this encounter, again including pneumothorax and pneumonia, and hydrothorax, empyema, and mediastinitis in various patients, with 1 nonpulmonary complication of an esophageal perforation at the gastroesophageal junction. None of these complications were prevented by the use of a cuffed endotracheal tube or tracheostomy tube. This led the authors to update their institutional policies, emphasizing the importance of confirming tube position by x-ray before starting a tube feeding, adequate training of all personnel involved with tube placement, and limiting the use of stiffening stylettes to the level of the nasopharynx only.8 Various novel attempts to minimize inadvertent malposition at the bedside include the use of an attached "stethotube" by Haddad et al.9 The presence of audible breath sounds leads to prompt NG tube removal and reattempt at proper insertion. A new technical development by the Viasys Company (Conshohoken, PA) includes a Cortrak insertion system, involving use of a "global positioning system"–type technology adapting an unweighted nasoenteric feeding tube with a copper wire, using an external transmitter with an electromagnetic pulse sensor system at the bedside to observe the location of the tip of the tube as it is manually inserted. Preliminary studies imply success of intubation, avoiding lung placement in 100% of cases studied, although early studies are somewhat small in number.10 This system has the potential of avoiding inadvertent tracheal or right or left mainstem bronchus insertions before inadvertent lung puncture and subsequent pleural-space insertion. This has the potential for decreased insertion time to the stomach and potentially even duodenum or jejunum placement, although this has not yet been thoroughly studied. A recent study by Bercik et al11 studied a magnet-tipped computer tracking system for tube placement in normal volunteers and compared it with simultaneous fluoroscopic and manometric monitoring. Excellent 100% correlation was noted between the 3 techniques. The authors felt this may be a useful tool to aid in bedside tube placement, at least into the stomach.11 New commercially available carbon dioxide (CO2) sensor devices can be attached to a feeding tube to assist in proper esophageal and gastric placement. These colorimetric devices rapidly change color if CO2 is present, implying inadvertent lung placement, and allow repositioning the tube without additional unnecessary x-rays. Authors reporting cases of inadvertent pleural access stress the awareness of this relatively rare but potentially life-threatening complication, as tension pneumothorax in a compromised critical care patient could be catastrophic. Prompt recognition, removal of the tube, urgent radiograph of the chest, and even potential rapid bedside chest tube insertion may be necessary if tension pneumothorax develops due to the adverse outcome of transbronchial nasoenteric tube perforation of the pleural space. Subsequent pleural effusions or empyemas would require appropriate surgical management with drainage and antibiotics. Most institutions now insist on routine chest x-rays after all bedside blind nasoenteric tube placements to prevent tube malposition that can lead to the above-noted pulmonary complications.7,8 Many institutions now require x-ray confirmation of NG tube in the stomach rather than the esophagus before infusion of bowel preparation solutions, medications, or tube feeding to avoid inadvertent fluid or tube-feeding aspiration.
Epistaxis/Sinusitis Sinusitis has been reported, more frequently as a longer-term complication after extended nasoenteric tube placement. Incidence is variable, ranging 11.4%–13% when diagnosis is based on aspiration and culture of the frontal and ethmoid sinuses, to as high as 25% when diagnosed by x-ray alone, which tends to overestimate the occurrence according to a study by George et al.13 An additional otolaryngologic complication involves patients with midface fractures, where inadvertent NG tube placement has entered the cranial vault. This rare complication was reported by Genu et al.14 Patients requiring long-term ventilator support with potential increased risk of sinusitis from nasal intubation may benefit from an orally placed feeding tube.
Inadvertent Tube Removal
Tube Clogging A clinical pharmacist can be helpful, ensuring that medications are appropriately crushed and flushed and that drug-drug interactions are kept to the minimum. A study by Powell et al17 showed a significant decrease up to 10-fold (from 66% to 7.6%, p < .05) in tube clogging with a protocol change, eliminating gastric residual aspiration. The author surmised that gastric acid aspirate into the tube contributed to the clot/clog formation. Additional protocol options to minimize tube clogging can include larger tube size (10 Fr vs 8 Fr) and possible use of semielemental vs standard or fiber-containing formulas. Marcuard et al18 showed that if one attempts to declog a tube with a chemical agent, a pancreatic enzyme suspension mixed with bicarbonate to optimize pH was superior to nonalkalinized pancreatic enzyme and commercial carbonated beverages. If the tube clog cannot be adequately managed with aggressive water flush or the above alkalinized pancreatic enzyme regimen, then mechanical clearing can frequently be successful with the use of various endoscopic catheters, braided guidewires, endoscopic cytology brushes, or commercial "cork-screw devices." Care must be taken to ensure that the length of the clearing device has been measured against a similar-length feeding tube to avoid inadvertent intestinal perforation of the duodenum that could occur with an excessively long device.19 In addition to clotting and clogging, tubes left in for an excessive period of time can have biologic deterioration, sometimes associated with Candida invasion of the polymers, discussed in more detail under percutaneous endoscopic gastrostomy (PEG) tube malfunction. Occasional NG and nasoduodenal tubes have been confirmed by x-rays and fluoroscopy to have actually tied themselves into a knot, leading to tube occlusion and difficulty in manual removal.
Tube-Feeding-Associated Diarrhea
Aspiration Pneumonia After reviewing these and other articles on this controversial topic, aspiration is clearly associated with long-term nasoenteric tube feeding, ranging from 25% to 40% incidence.24 However, trying to establish cause and effect has been much more difficult. As Heyland and Parker22 pointed out in their recent article, current concepts for ventilator-associated pneumonia (VAP) seem to implicate aspirated oropharyngeal or gastroesophageal contents that may have been colonized, with colonization altered by patient sensorium, pH of gastric juice, the presence or absence of concurrent antibiotics, and presence of contamination in the mouth. Many of these factors may actually have nothing to do with the actual enteral nutrition product or delivery system.22 Various preventive techniques were discussed, including the use of Chlorhexidine to the oropharynx to modify bacterial colonization25 and the use of a 45-degree elevation for semirecumbency.26,27 Drakulovic et al28 documented increased mortality associated with aspiration pneumonia in patients randomized to the supine vs semiupright position. Another controversial but potentially important new area of potential preventative treatment in VAP is the use of selective digestive tract decontamination (SDD).29 There have been at least 7 meta-analyses evaluating over 40 randomized, controlled trials of the use of SDD. The trials usually include a regimen of topical antimicrobials with polymyxin, aminoglycoside, Amphotericin B and sometimes a short course of 3–4 days of systemic antibiotics. These protocols all support the concept of eradicating early tracheal colonizers such as Streptococcus pneumoniae and attempt to prevent later colonization of the GI tract and trachea with aerobic gram-negative enteric bacteria. All 7 published meta-analyses showed a significantly reduced risk of ventilator-associated pneumonia, with 4 of 7 also showing a significant reduction and mortality.29–32 The primary caution of this therapy however revolves around the possibility of inducing resistant organisms that would complicate future management of any potential "super infection."
Use of Blue Dye
Postpyloric Tube Placement: Endoscopic vs Fluoroscopic Techniques Successful postpyloric placement can be achieved by numerous techniques, but frequently is complicated by difficulty in deep enteral access, usually defined as insertion to the ligament of Treitz or the proximal jejunum. A landmark research study in normal volunteers by Gutske et al36 studied high infusion rates of dye markers at various positions, ranging from the duodenal bulb to the true jejunum beyond the ligament of Treitz. Even at rates significantly higher than that in clinical use, position in the fourth portion of the duodenum or proximal jejunum led to negligible duodenal to gastric reflux, presumably showing safety of distal duodenal and proximal jejunal feeding in minimizing small bowel regurgitation as a component of potential aspiration.36 In an early landmark study by Zaloga,37 bedside placement using a special technique that included the use of a stiffening stylette and a "corkscrew technique" placed 92% of the tubes in a postpyloric position, although only 64% were in the distal duodenum or proximal jejunum. Procedure time averaged 40 minutes, decreasing to 10–20 minutes after additional experience. This was a modification of a previously described technique by Thurlow.38 Although this bedside nonendoscopic nonfluoroscopic technique has achieved significant success in numerous studies, it is significantly operator-dependent, has a significant learning curve for an optimal success rate, and it is not readily available in most community hospitals who do not have a designated individual or "team" of clinicians trained in this specialized insertion technique. Due to this lack of availability and quality control, various endoscopic, fluoroscopic, and combined endoscopic fluoroscopic techniques have been described. Bosco et al16 described a through-the-scope technique, using a large therapeutic scope 8-Fr feeding tube and Teflon-coated guidewire-assist technique. They achieved a 90% success rate of proximal jejunal placement in 10 patients, confirmed by subsequent x-ray. Average procedure time was 19 minutes. Baskin and Johanson39 described a combined endoscopic and fluoroscopic technique, with bedside fluoroscopy to assist in insertion speed. They used a combined double-lumen tube in which one lumen of the tube was an 18-Fr NG suction and the other lumen was a 9-Fr distal NJ extension tube. The tube was place using an oral endoscopy technique with nasal backload of a guidewire, followed by fluoroscopic placement. Mean procedure time was 20 minutes in their initial series of 37 patients, decreasing to <10 minutes after additional experience was gained. Success rate was 100% for jejunal access, using this combined technique.39 Patrick et al12 subsequently had a 94% success rate of jejunal placement with the same double-lumen tube, with modification of the endoscopic technique using an extended guidewire and grasping forceps and guidewire tension, allowing distal placement without the need for fluoroscopy. Their average procedure time was 12 minutes. The primary technical difficulty with these techniques revolves around the problem of use of an oral standard endoscope, which requires nasal back-loading using a transfer tube to subsequently allow the NJ feeding tube placement. This can be awkward, including inadvertent wire looping around the uvula, patient biting of the operator's finger during finger-assisted wire straightening, and periodic excessive epistaxis and subsequent sinusitis from prolonged use of the large-diameter NG suction tubes. However, successful start of enteral nutrition frequently leads to resolution of any clinical ileus, rapid cessation of high gastric outputs, and the potential for exchanging the double-lumen tube to a single-lumen 10-Fr tube over a guidewire with bedside fluoroscopy, potentially avoiding some of the long-term nasal complications.39 More recently, transnasal endoscopy has been proposed to eliminate the nasal back-loading step, allowing rapid bedside placement with full endoscopic control, with the use of the 5- to 6-mm "mini" pediatric endoscopes currently available. Detailed discussion of the various techniques used to carry out this procedure have been described by Dranoff et al,40 Kulling et al,41 and a review article by DiSario et al.42 The role of fluoroscopy remains somewhat controversial. It clearly is a successful procedure done usually by a radiologist but unfortunately usually requires transport of critically ill patients, frequently over long distances, to the radiology suite. This has the added problems of cost and x-ray exposure. Numerous case reports note potential problems with loss of venous and arterial lines and minor patient injuries associated with x-ray table transfer.37 When available, bedside fluoroscopy with a portable "C-Arm" unit eliminates the patient transfer complications. An interesting randomized trial of endoscopic bedside vs fluoroscopic placement of postpyloric feeding tubes was carried out by Foote et al.43 Their results showed no significant difference in either the success rate or time of placement between endoscopic or fluoroscopic placement of feeding tubes. They did not, however, carry out a cost analysis of these 2 relatively expensive procedures.
Residual Volume
A classic study by McClave et
al44 found little
correlation with adverse clinical outcomes with progressively higher residual
volumes and concluded that Welch et al47 studied the validity of 4 indicators to try to predict successful duodenal feeding tube placement in a prospective trial. The authors looked at auscultation, vacuum effect, change in pH, and change in color, and concluded that a positive auscultation change or vacuum effect test was not conclusive for duodenal placement. A positive pH or color change may be significant and obviate the need for a confirmatory radiograph. However, nearly 40% of these "positive" test patients had duodenal location by x-ray in the proximal duodenum, and they felt further studies were warranted before elimination of confirmatory radiographs. A more recent study by Phang et al,48 looking specifically at pH measurements of tube-feeding aspirates to differentiate gastric vs enteric location, concluded that pH value was a reliable predictor of nasal duodenal tube placement. However, they were unable to exclude the need for radiographic confirmation due to the low sensitivity of pH measurements alone. A pH increase of >1.0 was at least strongly suggestive of duodenal placement.
Clearly, the introduction of the PEG procedure initially described by Gauderer et al49 more than 25 years ago certainly began a revolution in percutaneous enteral access, adding a safe, successful route for enteral nutrition in patients previously unable to be easily fed.49,50 Since that time, there has been a near exponential rise in the use of this technique, with an estimated 216,000 procedures performed annually today.51 Early PEG kits were frequently homemade, splicing together Pezzar catheters and angiocatheters with suture material and hand-tied knots. This led to periodic complications of inadvertent separation of the tube and the pull suture, occasionally with catastrophic results. Currently, manufacturers have markedly improved commercial PEG kits, increasing the insertion length, minimizing the risk of premature separation of the dilating obturator catheter, and modifying the internal bumper or bolster for ease of manual removal and minimizing buried bumper syndrome (see subsequent discussion on buried bumper syndrome). Table 1 illustrates features of an ideal PEG kit, which can be used in clinician purchase decisions. Although indications for PEG placement have expanded in recent years, frequently including any patient with malnutrition and the need for long-term enteral nutrition access, specific indications where PEG placement is ideally suited are listed in Table 2. Specific contraindications are also listed, although some of these are now considered relative, and special techniques in dealing with these relative contraindications are to be discussed subsequently.
Although PEG tube placement is considered safe and efficacious and is generally perceived as a "low-risk procedure," numerous complications have been reported over the 25-year history of this procedure, ranging from rare to common, with occasional catastrophic results. These potential complications will be discussed sequentially, again reviewing their incidence, potential causative factors, recommended management, and potential preventative measures when available.6,50,51
Procedure-Related Complications More significant procedure-related complications include inadvertent puncture of the colon, small bowel, and liver. Colon and small bowel puncture may be unappreciated at the time of the insertion as the PEG can be pulled through both lumens, causing apparent adequate gastric access. When the tube is removed, reinsertion of a replacement PEG, particularly of the manually inserted silicone balloon type, may lead to tube feeding infusing into the small bowel or colon. Subsequent lack of appropriate absorption and rapid-onset diarrhea may require contrast, computed tomography (CT) scan, or endoscopy or colonoscopy for diagnosis. Numerous case reports of this complication, known as gastrocolocutaneous fistula, have been reported.57,58 This relatively rare but potentially serious complication can be prevented with a specialized technique popularized by Foutch,59 known as the "safe-tract" technique. This requires the use of a small-gauge aspirating syringe filled with water or saline. The assistant can puncture the stomach under direct endoscopic vision, watching for the simultaneous needle puncture into the gastric lumen, with presence of air bubbles. Intervening air bubbles imply perforation of a hollow viscous, usually transverse colon, less likely small bowel, which implies the tract is not "safe." Care must be taken, however, that the assistant follows the exact oblique angle of the safe-tract needle when the sheathed angiocatheter is punctured, as a "pseudosafe tract" can be bypassed if the angle of the needle is altered, particularly in a more superior inferior position rather than an oblique insertion position.59 Graphic demonstration on how the safe-tract technique can prevent inadvertent colonic puncture during initial PEG insertion procedure is described in Figure 1.
Much more serious inadvertent puncture of the left lobe of the liver during PEG insertion, though quite rare, has been described several times, including case reports by Chaer et al60 and Gubler et al.61 Of interest, all 3 cases described in these references were asymptomatic, found during radiographic evaluation for unrelated causes. As the tract remained stable and it was felt that removal may have been hazardous, each of these cases was managed nonoperatively, leaving the PEG site intact, without any short-term complications reported; specifically, no abscesses were noted. Procedure-related bleeding, related to the small skin cut done usually for both pull and push techniques, is usually considered minor and resolves spontaneously.52 A more recent study by Sedlack et al62 studied 20 consecutive patients with 20-Fr PEG tube placement without any skin incision. They had no infectious or bleeding complications but did note that mean pull forces required for placement were greater than that published previously for PEGs with skin incision. An accompanying editorial by Ponsky63 regarding this paper noted lack of a control group and a higher incidence of tube failure. He felt that a randomized comparison of the 2 methods would be warranted to see if this truly would decrease the small risk of procedure-related bleeding. Prolonged ileus can occur in up to 3% of cases, with occasional acute gastric dilation reported, and can be managed with venting the PEG tube to dependent drainage and holding tube feeding for a period of time.64
PEG-Site Infection Prospective, randomized trials have shown improvement in this variable incidence of site infection, with administration of a single broad-spectrum antibiotic such as a third-generation cephalosporin.67 In addition, authors have recommended that cleansing the patient's oral cavity with cyproheptadine or other antiseptic mouthwashes should be done prior to PEG tube placement to minimize the pull-through effect of mouth flora contaminating the skin-cut PEG site.67 One of the contributing factors to apparent skin-site infection and breakdown is excessive tension of the external bolster, frequently with the addition of numerous layers of gauze dressings. Removing the gauze and loosening the external bolster, allowing the site to air dry with minimal external tension, can prevent or treat local site infection, particularly if systemic antibiotics have already been started. Incision and drainage by a surgeon can be done if a clear-cut abscess is present but is not usually necessary.6 An example of proper PEG placement, with light application of external bolster, with or without a minimum amount of gauze between external bolster and abdominal wall, is illustrated in Figure 2.
PEG-Site Leakage A not-uncommon complication after PEG placement is that of persistent leakage or seepage of either tube-feeding formula or gastric juice. Progressive excoriation of the skin around the site of the tube can at times represent a second-degree burn and can evolve into a medical emergency. Although the reported incidence is considered low at 1%–2%, most clinicians in the community setting feel this is more common than is reported.68 Numerous contributing factors have been reported, including excessive cleansing with hydrogen peroxide at the skin site, excessive torsion along the PEG tube with an extra tight external bolster, absence of gastric acid suppressive medication, and general poor nutrition of the patient not yet corrected. Medications may need to be reviewed and proton-pump inhibitor therapy started if not currently being used. Side torsion of the tube should be avoided, particularly if the tube is used for dependent drainage. The enterostomal nurse clinician should be consulted for appropriate antifungal cream or zinc oxide application and potentially even a commercial ostomy wafer with attached tube clip holder and external ostomy bag to control the drainage.
Potential conversion of the PEG to a PEG/J double-lumen system, diverting tube feeding downstream, has also been reported as beneficial.6,50
Buried Bumper Syndrome Incidence of this complication is also highly variable, ranging 0.3%–2.4% of patients.69 In the usual clinical presentation, excessive pain is noted at the PEG site. Most important, the bedside nurse or the patient notes that the PEG tube can no longer be gently inserted and twisted or rotated with ease and appears to be fused to the abdominal wall. Tube-feeding formula, medications, or contrast can no longer be instilled into the gastric lumen. This is usually confirmed by endoscopy, noting complete or partial absence of the internal bolster endoscopically.6 Numerous novel techniques have been reported in terms of management of buried bumper syndrome, with the goal of all of these procedures in some way to remove the buried device and either reaccess the luminal tract with a new tube or, when necessary, secure an entire new access site. Techniques have included the insertion of a second push-type PEG to push out the buried bumper, use of an endoscopic needle knife internally to allow the embedded bumper to be reinserted into position, or the use of percutaneous balloon devices to expand the lumen, allowing external removal and pull-through replacement.19,70–72 Detailed discussion of the techniques used for potential buried bumper removal are described in detail and summarized in a review article by McClave and Chang.6
PEG-Tube Malfunction Despite the current pharmacy recommendations that each medication be separately flushed to avoid drug-drug interactions, bedside caregivers both professional and amateur have been observed to instill medication sequentially without flushing, potentially contributing to this dysfunction. Occasionally, devices such as biopsy forceps, endoscopic guidewires, and commercial plastic "decloggers" or brush-type devices have been used to clear an inspissated tube lumen, allowing it to be used again and avoiding the expense of tube replacement.73 Petroleum-based polymers, used in the manufacturing of various PEG tube devices, have been shown to be degraded by fungi, leading to actual colonization with Torulopsis, Candida, and Aspergillus species. These characteristically leave spots or stains growing from the inside of the PEG lumen to the outside, leading to rapid deterioration of the PEG tube. In compromised patients, including the elderly and the diabetic, source this is a potential of fungal sepsis.74–76 Tubes with this apparent complication, including significant deterioration of the material with cracking and splitting or evidence of fungal colonization, should obviously be removed and replaced if still needed for enteral nutrition.
Inadvertent Removal
Miscellaneous PEG Complications The presumptive mechanism is direct seeding of tumor cells spread by traumatic push- or pull-through of the PEG bolster as it is maneuvered through the head and neck tumor, usually in the pharynx or proximal esophagus. One case report of palliative radiotherapy has been noted in the clinical literature.79 Some authors, concerned that this rare but potentially serious complication could compromise the potential cure of a patient with head- and neck-cancer, recommended the radiographic Russel introducer-type technique, avoiding pushing or pulling bolsters through the tumor.6 An additional potential clinical problem revolves around PEG placement in patients with ascites. Various techniques have been tried to manage this problem successfully. Kynci et al80 carried out large-volume ultrasound-guided paracentesis and use of diuretics, albumin, and Mannitol, before PEG. Venous Doppler ultrasound marked potential abdominal varices, and a PEG tube was placed successfully. More recently, Wejda et al81 described a new technique involving gastropexy with a somewhat complicated system of "T fasteners" and sutures to avoid premature separation of the PEG from the abdominal wall, with reaccumulation of ascites. Two somewhat related PEG-insertion problems include PEG placement with ventricular peritoneal (VP) shunts and considerations for PEG placement in patients with ambulatory peritoneal dialysis catheters. Numerous authors have reported on the safety of PEG placement with VP shunt, and initially prophylactic antibiotics were not felt to be necessary. Subsequently, however, case studies of central nervous system infection have been reported, and major antibiotic prophylaxis is now recommended.82 In a related situation, most authors have felt that patients with healed PEG tracts can have subsequent ambulatory peritoneal dialysis catheters placed safely, with only antibiotic prophylaxis given in similar fashion to those patients with VP shunt placement. However, some recent case studies where PEGs were placed subsequently after preexisting ambulatory peritoneal dialysis was in place have shown a surprisingly high infection rate, including a moderate mortality. The authors conclude that the high dextrose concentration used with ambulatory peritoneal dialysis represents a significant culture media, and the frequent concurrent malnutrition and low serum albumin levels associated with patients with chronic renal failure make this an especially high-risk group. The authors concluded that acute hemodialysis should be a consideration before, during, and after PEG placement, only reverting to ambulatory peritoneal dialysis when the PEG tract was felt to be fully healed, assuming that ambulatory peritoneal dialysis was still an option.83 An extremely rare and unique complication of PEG placement has been reported by Rosemore et al,84 namely, temporomandibular joint dislocation during the endoscopic procedure. Five previous cases had been reported with standard endoscopy and bronchoscopy. This was apparently the first case associated with PEG-tube placement. The authors described the appropriate reduction of the dislocation and postulate that excessively long endoscopic procedure time and excessively large mouth pieces can be contributing factors, and recommended short procedure length and small mouthpiece size. Finally, patients with malignant dysphagia, tracheoesophageal fistulae, and saliva management issues periodically require a self-expanding metal stent. Frequently, authors recommended PEG placement before stent placement to avoid the theoretical concern of stent migration with push- or pull-through technique for PEG after stent placement. Adler et al,85 however, reported a retrospective review of 9 consecutive patients with indwelling esophageal self-expanding metal stents undergoing subsequent attempt at PEG. This was successful in all patients, with only 1 case of distal stent migration that was managed endoscopically without further complications. The authors concluded that this was a relatively safe and feasible procedure, recognizing the potential for distal stent migration, and recommended managing that endoscopically in the same setting. Current recommended complication-management guidelines are summarized in Table 3, including procedure-related complications, postprocedure complications, and prevention and management suggestions. These guidelines have been provided by Lynch and Fang.51
One of the more technically complicated, frustrating, but frequently attempted nutrition support procedures is that of PEG/J placement. This new terminology was the outcome of an American Society for Gastrointestinal Endoscopy (ASGE) task force on enteral nutrition, which recognized that the numerous abbreviations, names, and descriptions of this procedure could best be consolidated into PEG/J, particularly to avoid confusion with the newer and more accurately described direct PEJ (DPEJ), to be described subsequently.42 The apparent reason for the multiple attempts at perfecting this procedure stems from the fact that early on, patients are frequently intolerant of gastric feedings, and even after a successful PEG placement, high gastric residuals, apparent reflux and aspiration, bloating, abdominal distention, and nausea frequently persist.50 The clinician assumes that if a feeding tube could be passed beyond the pylorus, accomplishing the goal of postpyloric feeding concurrent with gastric drainage to a dependent drainage bag, these complications could be avoided. Unfortunately, early "drag and shove" techniques, initially described by Ponsky and Aszodi86 in 1984, were universally unsuccessful in accomplishing this goal. Numerous newer techniques have been described, with varying degrees of success, including a classic guidewire-assisted technique described by Duckworth et al,87 a steerable guidewire technique modified by Parasher et al,88 and a unique PEG lumen snare with retrograde wire pull-through assisted technique described by Leichus et al.89 Subsequently, Baskin and Johanson90 further modified the steerable guidewire technique with the use of an ultrathin "pediatric" miniscope passed directly through the PEG lumen, requiring a 28-Fr PEG tube but allowing a 12-Fr PEJ tube. Figure 3 illustrates this easy-to-learn successful technique. Detailed descriptions of these various PEG/J techniques were described previously in a review article by Baskin.50
Unfortunately, the most common complication of this potentially beneficial procedure includes periodic retrograde migration of the jejunal extension tube back into the stomach and, in one case, even into the hypopharynx, leading to subsequent aspiration.91,92 Although a recent review by Guidroz and Chaudhary93 described the improved delivery of nutrition in trauma patients with PEG/J as compared with those with PEG feedings, numerous authors continue to criticize this technique, describing frequent tube clogging, positional kinking, medication occlusions, spontaneous PEJ knotting, and the aforementioned technical difficulties with achieving deep jejunal access, even using some of the newer techniques described above.94,95 Additional attempts at stabilizing the tip of the jejunal extension tube to avoid proximal migration have been described, using an anchoring segment of 1-cm tubing spliced to the PEJ tip with a 20-cm-long silk suture extension as a downstream weight.6 Ginsberg et al96 used an endoscopic clip fixing device, attaching the tip to the small intestine in 6 patients with apparent success. DeLegge and Kirby,97 however, in a subsequent editorial response felt that these techniques were unnecessary if NG/NJ and PEG/J systems had been placed properly with guidewire assist and fluoroscopic or x-ray confirmation, usually staying in place for an extended period of time, with no need for clip placement. Continued frustration with this technique by numerous clinicians, perhaps not fully versed on the breakthrough of the more successful PEG/J techniques described above, has led to much wider use of the new DPEJ technique initially described by Shike et al.98
Shike and colleagues98 described an entirely new technique for true "direct percutaneous endoscopic jejunostomy," now known as DPEJ. This is a definitely a more complex and higher-risk technique than the various PEG/J transpyloric extension tube techniques described previously: it has a steeper learning curve and a new set of complications. Numerous recent studies, however, attest to its potential benefit, particularly in long-term management of patients that cannot otherwise be safely fed.98–101 The specific placement technique used by Shike and others98 is nicely diagrammatically outlined in a review article by Ginsburg,99 summarized in Figure 4. Rather than a 1-doctor, 1-nurse procedure, as noted in Figure 4A, this usually requires 2 nurse assistants and 2 physicians to carry out this more complex technique. To avoid some of the technical difficulties and complications with some of the commercial styletted introducers, in order for adequate maintenance of small bowel access due to gut motility, a modification of this technique was described by Varadarajulu and DeLegge.100 They used a small-gauge safe-tract-type needle, grasping it with an endoscopic snare to maintain access, and then made a second puncture with a sheathed trocar, releasing the safe-tract needle and grasping the introducer trocar for safe guidewire pullout. They achieved a 92.3% success with this technique in 26 patients, with only 1 inadvertent small bowel perforation. Average procedure time was 23.3 minutes. Indications for DPEJ are listed in Table 4.98 Not infrequently, complicated prior esophageal, gastric, or pancreatic resections or the newer laparoscopically assisted gastric bypass procedures for obesity have led to impossibility of standard PEG placement, and DPEJ becomes a much more feasible technique in these patients with significantly altered anatomy.98
Complications unique to this procedure include inadvertent small bowel puncture at an alternate site to that of deep PEJ access, wound infections, subsequent abscesses, intraperitoneal bleeding, and colonic perforation. The success rate for this procedure is frequently less than that for PEG/J, particularly early in the learning curve. Success ranges from 86% to 92%, reported in large series with experienced operators.100,101 Shike et al98 reported a 6% incidence of incisional infections, managed medically. They had 1 each of bleeding, abscess, and colonic perforation, all requiring surgical intervention. They noted, in long-term follow-up, 3% incidence of tube malfunction and 3% incidence of aspiration, with a mean duration of tube use of 113 ± 73 days. A direct comparison study of DPEJ vs PEG/J was carried out in a study of over 100 patients. Using parallel survivor curves, DPEJ maintained patency at >90% for 6 months, with only 55% patency for PEG/J patients, a highly significant difference (p < .0001).102 Due to the potential efficacy for long-term management, a specific group of patients with documented severe chronic pancreatitis was studied by Stanga et al,103 using a combination of PEG/J and DPEJ. They noted significant increases in patients' body weight and decreases in the degree of malnutrition, abdominal pain, and associated gastrointestinal symptoms with long-term jejunal feeding and concluded that it was safe with minimal major complications. One interesting unusual complication of DPEG was recently reported by Chong and DeLegge,104 namely, necrotizing fasciitis after DPEJ. This apparently is the first reported case of this complication, although it is has been previously reported with standard PEG placement. The patient developed significant necrotizing fasciitis confirmed by CT scan, requiring subsequent surgical debridement. He unfortunately died 9 days later, despite aggressive surgical debridement and extensive antibiotics. It was attributed to peristomal leakage at the skin site and deep tissue seeding with bacteria. Prophylactic antibiotics did not prevent this infection. The authors recommended for any future cases early and aggressive surgical intervention, broad-spectrum antibiotics, and supportive intensive care unit care, given the potential high mortality risk.104
DPEJ Removal/Replacement Of interest, balloon replacement devices through jejunostomy tracts have been occasionally implicated in high-grade small bowel obstruction, with pressure necrosis of the small bowel and abscess, perforation, and death.105 Evaluation of these cases has shown that inadvertent installation of water flushes or medications through the Foley balloon port have led to massive distention of the balloon, with subsequent obstruction and the potential for ischemic necrosis and perforation.105
Radiologic Percutaneous Gastrostomy/Jejunostomy Reported complications of radiologically placed gastrostomy and jejunostomy tubes include inadvertent gastric vein puncture leading to bleeding, inadvertent percutaneous puncture of colon and liver (as described previously for PEG), leakage of gastric contents into the peritoneal cavity, periodic tube dislodgement, regurgitation of catheter into the esophagus, and wound infections. Minor complications have been reported from 3.2% to 6.9%, with major complications averaging 2.2%, with rare mortality (0.3%–0.5%).109,110
In summary, enteral nutrition continues to be an important category of management and is widely used in clinical practice. Unfortunate misadventures can occur, sometimes due to operator inexperience, inappropriate device selection, inappropriate patient selection, or unforeseen therapeutic misadventure. Many of the potential tube-placement complications outlined are preventable, but all have management possibilities. Clinicians need to be fully aware of the diverse potential complications that can ensue and be well versed on the number of different techniques that can be carried out to optimally achieve patient access appropriate for the patient's clinical condition. Hopefully, continued research into improved devices, access techniques, and patient and device selection criteria will improve the quality of patient care.
Nutrition in Clinical Practice, Vol. 21, No. 1,
40-55 (2006) This article has been cited by other articles:
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Top
Nasoenteric Tube Placement
PEG Tubes
200 mL would potentially raise concern, but
feeding can be continued while the patient is closely monitored. More recent
consensus
discussions
